Author name: Di Wu, Siyue Liu, Zixiang Ji, Ya-Liang Chang, Zhe-Yu Liu, Andrew Pleffer, Kai-Wei Chang

arXiv:2604.01354v1 Announce Type: new
Abstract: Moderation layers are increasingly a core component of many products built on user- or model-generated content. However, drafting and maintaining domain-specific safety policies remains costly. We presen…